Data on Indian Americans: religion & politics

Unfortunately there’s not as much quantitative data on Indian Americans as I’d like. To be fair, you can say that about almost anything, by which I mean there’s always a lack of enough data for my taste. One good source is the Religious Landscape Survey. 90% of Indian Americans are not Hindu, but 90% of Hindu Americans seem to be Indian. But that’s suboptimal. The Census has some good information, but is moderately constrained in what it can give you. There is one option which I’ve avoided for a while, the General Social Survey. This is a huge database which you can query if you are comfortable with using forms on the web (you should be). You can limit to people who say that their ancestors are from “India.” Unfortunately the sample size is only in the hundreds.

But I thought I’d give in a try, because I wanted to look at some intra-community cross-tabs. The main aspect I wanted to look at is religion & politics. Indian Americans are overwhelmingly a Democratic-leaning community. But not all. This tendency toward the Democrats has been relatively strong in Asian Americans generally since 1992, when George H. W. Bush won that demographic. And yet I noticed an interesting trend in the American Religious Identification Survey 2008: Asian American Christians were far more sympathetic to Republicans than Asian American non-Christians. The past 20 years has seen a massive rise in the proportion of non-Christian Asians, whether it be Hindu, Sikh, Muslim, Buddhist, or secular. The standard narrative in American politics is that the Republican party is the white Christian party (even more so, the white Protestant party). The Democrats are the coalition of “Others”. Minorities and non-Christian whites (seculars and Jews). This has clear first approximation value, but I think the insight that Asian American Christians are more sympathetic to Republicans than other Asian Americans indicates that there is some texture which can be perceived at a finer-grain.

My goal here is exploratory, and I want to encourage readers to poke around the GSS themselves. In short I limited the data set to 1990 and later, to people who said their ancestry was from India. Unfortunately this is only a few hundred, but it may be informative for large between class differences. I focused on differences across religion and levels of education. Some notes:- I combined Christian denominations as well as Dharmic religions to maximize sample size.

  • For the education oriented cross-tabs I combined those without college degrees and those with college or greater into two classes.

  • I dropped Muslims because the sample size was too small.

  • The columns add up to 100%. The percentages are bold. Underneath them you see the N.

I apologize for the formatting on the tables, but they’re screenshots of the GSS. The data is for real, even if the presentation is 1997.

Here’s a screenshot of the page I started with:

gss0.png

The numbers represent codes. As someone who’s used the GSS a lot I know that for the RELIG variable 1 = Protestant, 2 = Catholic, etc. Most of you probably don’t, so you might want to copy my starting position to begin with if you want to start cranking out tables.

First, let’s look at religion:

gss1.png

gss2.png

gss3.png

gss4.png

Now college education::

gss5.png

gss6.png

gss7.png

gss8.png

Interpretation? I think the truism for Asian Americans found in ARIS 2008 holds for Indian Americans: non-Christian Indian Americans are very averse to the Republican party and conservative ideological orientation. Looking at the religious identification by age cohort (not shown, use the COHORT variable) it seems pretty clear that because of high immigration levels the Indian American community is going through the same transition as the rest of the Asian American community: the proportion of Christians is decreasing. I suspect the shift away from Republican identification is then in part due to double alienation from the identity politics of the modern American Right, that of race and religion, common among immigrant heavy Asian American subcultures. To check, let’s look at those whose ancestors hail from Japan and China separated by those who are Christian vs. those who are not:

gss9.png

Again, the same tendency. This shouldn’t be that big of a surprise, after all the same division is mirrored in the general population. The Republican party and conservative orientation is associated with evangelical Protestant Christianity among whites, while Democratic leanings and liberal views tend to be associated with Jews and the irreligious (with Catholics and mainline Protestants somewhere in the middle). The valence though is somewhat different in Asian immigrant communities, which are religiously more fluid and pluralistic. Though Christian identity doesn’t seem to imply that Latinos and blacks are particularly favorable toward the American Right,* it does seem to have some predictive value among Asian Americans. It can naturally explain the much stronger tendency of Korean Americans to be Republican as opposed to Chinese and Japanese Americans. Evangelical Protestantism is much stronger among Korean Americans.

Why the difference? Identities are bundles of various factors. Race, ethnicity, religion, and class. The history of black Americans in the United States makes them somewhat sui generis. Latinos are still shifted toward the working class, and working class and poor people of color tend to vote Democrat no matter their religious identity (in fact, depending on how you identify working class, this is true for whites as well, see Red State, Blue State, Rich State, Poor State). In contrast, Asian Americans are very diverse when it comes to class. There is a recent finding in social science that cultural values tend to matter most when people have ascended somewhat higher up Maslow’s hierarchy of needs. In other words, the “culture wars” are a middle and upper class luxury. Non-economic identity variables are particularly predictive on the upper socioeconomic strata. I suspect then that what’s going on is that non-Christian Indian Americans are often identifying with Jews and secular Americans on cultural values, as part of the “Other” coalition. Christian Indian Americans with a conservative theological bent are more likely to identify with the white Christian party, because there’s a basis for common alignment. If the thesis that education and income free people up to become more ideological and culturally conscious, then Christian identification among Indian Americans should predict Republican party identification among the more educated segment.

The sample sizes are small, but the GSS results point in exactly this direction:

gss10.jpg

* Though there does seem to be a very modest effect of Protestant identity among Latinos making them more amenable to Republican party identification or conservative ideological orientation.

24 thoughts on “Data on Indian Americans: religion & politics

  1. I’m glad Indian Americans aren’t voting Republican, but it surprises me. What I see and hear is: Indians don’t like black people. In fact, many Indians are openly racist. Indians certainly don’t like gay people. They disapprove of open displays of sexuality. They believe in total obedience from their children. They have strong family ties, but minimal social conscience, i.e. would not seem to prefer large government programs. If this were a white person in America, it would scream … “conservative”!

    Perhaps this is all trumped by the perception that the Republican party is anti-brown and anti-immigrant. I can’t see any other reason.

    • i think you’re confusing indian with indian american. even if you correct for the reality that most indian americans are 1st gen, they’re not a perfect sample representation of people in india. also, you can query some of the attitudes you are ascribing in the GSS.

    • The biggest drivers to vote conservative amongst desis is that many desis are in labor-intensive work, such as being self-employed, or blue collar work. They resent welfare schemes for abusers of these programs, but they may benefit (but not abuse) some of these programs. So in many ways, I believe that they are a single issue and unsophisticated voting block.

      If desis thought things through, they should ask themselves: 1. Which party supports a mosque at Ground Zero? 2. Who was responsible for the liberating immigration act of ’65? 3. Which party SNIP SNIP SNIP? 4. Which party is for hate crime legislation? 5. Which party is more secular? 6. Which is a party of obstructionists, human capital squanderers, feudal lords, and sovereign hijackers (like the Desh lands), and which party believes in good stewardship of its people?

  2. Who was responsible for the liberating immigration act of ’65?

    dude, do you know anything? or do you just make stuff up and hope it flies?

    http://www.govtrack.us/congress/vote.xpd?vote=s1965-232

    http://www.govtrack.us/congress/vote.xpd?vote=h1965-125

    granted, the parties have changed a lot, but it’s really infuriating when people just make stuff up based on their ignorance. stop it. and you should be a little less patronizing too, seeing as how you don’t really know as much as you think.

  3. Hey, Razib, you’re a numbers guy — based on past experience, how long will it take before the comments are closed on this post? I suppose we could average the number of comments that get published before the thread closes, or we could do it by time. :P

    (And yes, I would have signed in to comment, but that shit’s not working [with wordpress] yet.)

  4. Hey, Razib, you’re a numbers guy — based on past experience, how long will it take before the comments are closed on this post?

    since i’m the one that closes, i can’t give a useful answer. 2-3 days sounds good. depends on the quality of the discussion. i try and follow the threads, and when the conversation bores/annoys me i’ll close. if it is engaging and of interest, i’d leave it beyond that point. i don’t close threads to be capricious, i just am of the belief that discussions have a life cycle, and their purpose is to exchange information, ideas, and viewpoints.

  5. Where are you getting the statistic that 90% of Indian Americans aren’t Hindu? That sounds wrong to me. If you look at Ch. 3, p. 49 of the religious landscape study it reports that 55% of immigrants from “south-central asia” are Hindu.

  6. It occurs to me that probably you meant, “Hindus are not 90% of Indian Americans”?

    • yes. the prose there is unfortunate, and the confusion is my error. note that by “hindu” i do not include sikhs and jains.

      Of the under-represented Hindus a highly disproportional number are low caste gujaratis (patels etc ).

      nice to know that some indian americans care to recall who is, and isn’t low caste! i guess you’re not a sellout and are staying true to old culture :-)

  7. Is there any data on the percentage of American Hindus/Buddhists are from Bangladesh/Pakistan?

    • P.S. Not just from the survey.

      Also, delete “are” from the previous comment.

  8. 90% of Indian Americans are not Hindu,

    I am sure Hindu Americans are under-represented in America but not by that much.

    I have read that of the ~2.8 million Indian-Americans some 200,000 are Jains, 500,000 are Sikhs and 600,000 are Christians making them by far the most disproportionately over-represented of India’s religious communities. Muslim Indians are probably even more under-represented than Hindus. Atheists and agnostics are very likely heavily over-represented.

    Of the under-represented Hindus a highly disproportional number are low caste gujaratis (patels etc ).

    In India the Muslims outnumber the Christians by 5 to 1 in the abrahamic category but among Indian-Americans the reverse is true. Likewise, in India the Sikhs and Jains are a tiny minority in the dharmic category, but in America they are a huge chunk of it.

    • Ramachandra: I am sure Hindu Americans are under-represented in America but not by that much. I have read that of the ~2.8 million Indian-Americans some 200,000 are Jains, 500,000 are Sikhs and 600,000 are Christians making them by far the most disproportionately over-represented of India’s religious communities. Muslim Indians are probably even more under-represented than Hindus. Atheists and agnostics are very likely heavily over-represented.

      According to the Indian census from 2001, the portion of the different faiths in India are as follows:

      Religion (%) Hindus 80.5 Muslims 13.4 Christians 2.3 Sikhs 1.9 Buddhists 0.8 Jains 0.4 Total * 100

      Based on this data and your data, which I trust is correct:
      Religion (% of total Indians in USA) OVER/UNDER REPRESENTED RELATIVE TO INDIA Hindu 43 UNDER REPRESENTED BY A FACTOR OF 0.53 Muslims ?? ?? Christians 21 OVER REPRESENTED BY A FACTOR OF 9.1 Sikhs 18 OVER REPRESENTED BY A FACTOR OF 9.5 Jains 7 OVER REPRESENTED BY A FACTOR OF 17.5

      By the way, I took the TOTAL value of Hindus, which may include Hindus whos origins are Nepali, Balinese, Trini/Tobago, Bangladeshi, etc. Also, there are a lot of quasi-Hindus, such as ISKCON (?), who do not identify as being Hindu. Anyways, for matters of simplicity, I assumed that all 1.2M Hindus were Indian which shouldn’t be a bad assumption.

      I have no clue why Christians, Sikhs, and Jains are over-represented, but my working theory is that these communities are enterprising (Jains and Sikhs are great commerce-oriented peoples) or have a proclivity to being expats (i.e. the Christians of South India who migrate as petrol engineers or laborers in the UAE).

      On another note, I don’t get the impression that Americans realize how over-represented Sikhs are in the USA. They imagine that much of India and Indian culture – or half of it – is Bollywood, bhangra, and turban culture, and that everyone is a “Poonjaabi”. They’ve never heard of a “Orissan” for example.

  9. Razib, I like reading about this sort of stuff (and your other Harappa project posts but one has so many graphs and numbers that I having a tough time following the text. Please tell me I am not the only one.

    • megha, sure, lots of data. but i invite you to poke around the GSS yourself.

  10. I attended the US Open table tennis tournament in Milwaukee, and there was no under-representation of Hindu Indians there at all! Neither of ethnic Indians settled in the US, nor of players directly from India, representing India in the competition. In fact, there was only one player out of the more than 50, who was of non-Hindu background, a Christian. Nice to see such a Hindu thronged and Hindu friendly event! Simply because, outside specific Indian community gatherings, it is so rare. I do know that in Canada, Sikhs and Moslems are over-represented and given way too much coverage by the media, in contrast to Hindus. Also, good to see Indians/person of Indian origin, participating hugely in a sport other than cricket!

  11. With Hindu Americans, most likely atheists and agnostics are a subset. Hindus do not have the same concept of being Hindu as christians have of being christian.

    Among the generation of Hindu Indians born here, I wonder what percentage of Hindus are actually agnostic even if they perform rituals with their parents and for their own weddings. As far as the generation of immigrant Indians, I do not see, strictly anecdotally of course, any difference in the atheist/agnostic makeup compared to Indians who remain in India. If anything, some may become even more religious after coming here as the temple serves as a cultural meeting ground too.

  12. the 80% figure includes the Dalits and I have seen a lot of their leaders question the 80% Hindu figure. I’d guess its about 60-70% tops

  13. With Hindu Americans, most likely atheists and agnostics are a subset. Hindus do not have the same concept of being Hindu as christians have of being christian.

    this is true, but a surprising number are believers in the supernatural agents.

    http://religions.pewforum.org/comparisons#

    click ‘belief in god.’ note the contrast with buddhists. the astika nastika divide?

  14. Razib, is there data on which groups comprise the remaining 10% of Hindu Americans? I’m guessing that a significant portion would still be of South Asian descent (Nepali, Bangladeshi, or from the Indian diaspora).

    I have actually known Balinese Hindus in the US, because there is a pretty great Balinese/Indonesian restaurant near where I live. But, I don’t know much about them, where they worship, etc. It would be nice if Hindu temples in the US could arrange for the different Hindu communities to meet and learn from each other as a celebration of diversity.

  15. http://religions.pewforum.org/portraits

    5%White (non-Hispanic) 1%Black (non-Hispanic) 88%Asian (non-Hispanic) 4%Other/Mixed (non-Hispanic) 2%Hispanic

    the sample size is REALLY small. but i find it entirely plausible that a substantial proportion of non-indian hindus are converts or ethnically atypical.

  16. That 90% of Indian Americans are not Hindu seems to me to be patently wrong! In my experience, a good majority of Indians, be it at work or at other social settings, are in fact Hindus. And I have been around quite a bit! I wonder how that figure came about!

    • that’s not what i meant. came out wrong. see above comments. i mean that less than 90% of indian americans are hindu.